Marie-Jeanne Lesot

DeltaXplainer clarifies model prediction differences for better user trust.

Innovative methods aim to reduce harmful online language while preserving original meaning.

A fresh perspective on how the order of arguments shapes discussions.